Figure 4 Zenodo CSVs

[sjspielman]

Towards #1692 ~At time of filing this is stacked on #1705, which is wrapping up checks for imminent merging so I expect this target will change to master before this is reviewed~ Amazingly, this happened about 10 seconds after filing 🚀

This PR does the Figure 4 CSVs; link to Figure 4. All of the figure 4 panels are in figures/scripts/, so no need to poke around in analyses/!

Some notes about each export:

Figures exported from figures/scripts/fig4-tp53-telomerase-panels.R

• 4A is an ROC curve. No wrangling needed for export
• 4B and 4C are violin plots which are handled by a shared function. I modified this function to also return the data frames that go into the plots, and then wrangle them a bit at the end for clean export.
  • Worth noting there is a summary DF that is also used to make these plots for ggpubr p-values and brackets, but I did not export it. All that summarized info can be calculated from the data that is exported, though!
• 4D export I had to make a Choice: Since the figure combines DNA and RNA results, we have both flavors of biospecimen ID. I ended up arranging on the RNA, not DNA.
• 4F is a forest plot. There is one additional forest plot CSV exported from figures/scripts/fig4-hgg-subtype-forest-plot.R. This export is basically a model summary.. feedback on what to export here appreciated!
  • It's column names term,name, and value aren't incredibly informative from an outside perspective, so I am thinking that a README.md in the Zenodo upload folder that defines what is meant by the columns could be the move here?

Exported from figures/scripts/fig4-heatmap.R

• 4E; I exported the data frame from which all matrix-y type inputs to heatmap functions were derived. This data is in a "wide" format and I did not tidy it, which was a Choice for consistency w/ the code. Any feedback?

Exported from figures/scripts/fig4-hgg-subtype-forest-plot.R

• 4G; same notes as for 4F

Exported from figures/scripts/fig4-hgg-kaplan-meier.R

• 4H; this is a KM curve... I ended up exporting the piece of data that I think has the most information from the model? Feedback here would be great too!

[sjspielman]

Noting that I want to start the zenodo-upload README before we get too far here, so another commit is forthcoming to this branch with that.

[sjspielman]

Noting that a partial README has been added, and it's all the stamina I have for it for this PR. It will continue to get expanded over the many zenodo CSV PRs!

[sjspielman]

Only code in figures/scripts/, which does not run through CI, was modified in this PR, so we can merge before checks are complete.